Approximately 20 million Americans and ten times that many persons worldwide have diabetes mellitus. This disease is a chronic disorder that requires careful regulation of glucose levels within tight limits in order to prevent severe secondary complications involving the patient's eyes, kidneys, nerves, and blood vessels. Recent research from our group has demonstrated that surface-enhanced Raman spectroscopy (SERS) is a new approach to this important public health problem that offers significant promise as a means to measure real-time, in vivo glucose levels. Many of the potential problems envisioned at the outset of this project such as: (1) glucose had never been measured by SERS; (2) temporal stability; (3) reversibility; (4) real-time response; (5) resistance to protein interference; and (6) complications from interfering small molecule analytes have been demonstrated to be non-existent or have been overcome. The research proposed herein is focused on the remaining fundamental scientific and technical challenges associated with developing in vivo SERS as robust, portable biosensor platform for glucose in biological fluids. Four specific aims to designed to achieve this goal are: (1) synthesize and optimize new partition layers that reduce spectral overlap with target analytes, increase sensitivity, and increase in vivo operating lifetime; (2) improve the chemometric data analysis and calibration methods used to extract the glucose level from the raw signal ; (3) develop both fiber optic and free-space excitation/collection approaches to transdermal SERS; and (4) use a rat model to quantify glucose levels in unrestrained, unanaesthetized animals for periods up to 5 days. New partition layers will be created through synthetic chemistry to minimize spectral overlap. The operating lifetime of the partition layer will be dramatically increased using a novel approach based on atomic layer deposition to eliminate the weak Ag-S or Au-S bond. Calibration will be improved by measuring the glucose partition coefficient using liquid chromatography and by using a robust internal Raman standard (diamond). Transdermal SERS studies will be carried out to demonstrate the viability of free-space laser excitation and collection. Preliminary results are presented indicating that this is possible. In vivo SERS will be cross-validated using an implanted electrochemical sensor (MiniMed) and ex vivo monitoring of glucose by liquid chromatography/mass spectrometry.